DE69222884T2 - Heterobifunctional antibodies with both catalytic properties and specific antigen binding properties and methods of using them - Google Patents

Abstract

A novel and useful class of antibody molecules is described which possess the characteristics of an antigen-specific antibody and a catalytic antibody. Such a hybrid molecule may be assembled in vitro using biochemical means or using in vivo methods including cell hybridization, or recombinant DNA technologies. The product of the invention possesses both antigen target specificity as well as an intrinsic catalytic activity. Further said product may be used in any appropriate application including diagnostic test as a substitute for an enzyme-labelled antibody conjugate or as a therapeutic reagent whereby a part of the antibody molecule recognizes a unique cellular marker and the catalytic portion of antibody molecule has therepeutic activity.

Description

Field of the invention

This invention relates to bifunctional antibodies having both catalytic activity and antigen binding specificity and to methods of using such antibodies.

Background of the invention

Polyclonal antibodies are known for their different specificity ranges.

In 1975, Kohler and Milstein published a strategy for immortalizing lymphocytic cell lines (i.e. hybridomas), which are characterized by their ability to produce monoclonal antibodies with unique antigen specificities. Since this development, countless hybridomas have been developed with specificities against an equally large number of antigens. Some examples of the types of molecules recognized by monoclonal antibodies include carbohydrates, amino acid sequences, nucleic acids, conformational features of molecules, and inorganic side groups.

Antibodies share certain properties with a catalytic class of biomolecules known as enzymes. Both antibodies and enzymes are protein-like, have a binding cleft for a corresponding ligand, and exhibit a high degree of specificity for their respective target molecules. The binding cleft of an enzyme provides an arrangement of functional groups to catalyze a specific chemical transformation of the bound ligands.

In 1975, Raso and Stollar attempted to generate antibodies that specifically bind analogues of the intermediate involved in enzymatic and non-enzymatic pyridoxal phosphate catalysis. Although they were successful in generating such antibodies, they did not report any catalytic activity (Raso J. and Stollar B.D., "The Antibody-Enzyme Analogy. Characterization of Antibodies to Phosphopyridoxyltyrosine Derivates", Biochemistry, vol. 14 (1975), 584-591).

Trannontano and Lemer used the hybridoma technique in the mid-1980s to select an antibody that could recognize the transition state form of a substrate. As described in US Patent 4,900,674, Tramontano et al. used a synthetic phosphonate analogue that matched the putative transition state form an amide or ester product to develop and identify monoclonal antibodies that recognized the phosphonate substrate analogue and coincidentally catalyzed the formation or cleavage of the corresponding carbonyl ester or amide.

Since then, other catalytic monoclonal antibodies have been developed. For example, antibodies with specific catalytic activities have been found, which include such activities as lactone cyclization, amide cleavage, and acyl group transfer.

Each of these catalytic antibodies was developed using an analogue similar to the putative transition state form of an antigen or hapten to develop and identify the desired antibody. In each case, antibodies were developed that specifically bound to the immunizing hapten or antigen, generating an antibody that had truly catalytic properties; see Lerner et al., Science 252 (1991), 659-667.

Studies show that such antibodies exhibit enzyme properties: the catalytic conversion follows the principles of Michaelis-Menton kinetics; the antibodies are not consumed during the conversion; the antibodies show a characteristic constant turnover rate and the conversion can be inhibited with the original immunogen or an analogue thereof.

The term "abzyme" was coined to designate an antibody molecule with such catalytic activity as an internal part of the molecule; see Jarda et al., Science 241(1988), 1188-1191. This term is used in this specification to describe such antibody molecules.

Advances in hybridoma technology have recently led to the development of another type of antibody molecule. US Patent 4,474,893 describes methods for producing antibodies that have binding affinities for two different antigens. Antibodies of this type have been obtained by fusing hybridomas, which secrete monoclonal antibodies with binding specificity for one antigen, with hybridomas or lymphocytes that produce antibodies with binding specificity for another antigen. The fused cells, referred to as quadroma or trioma, produce antibody molecules of which 10 to 50% have binding specificity for both antigens; see Reading C. in Hybridomas and Cellular Immortality, Plenum Press (1981), 235-250, and Milstein C. and Cuello A.C. in Immunology Today (1984), 29.

Antibodies with binding properties for two antigens were called heterobifunctional antibodies. Such antibodies were also obtained using chemically dissociated antibody halves followed by reassociation using chemical cross-linking agents.

Heterobifunctional antibodies have been developed by researchers attempting to develop a vehicle for drug delivery. In particular, it was considered that an antibody moiety specific for a target cell paired with an antibody moiety specific for a toxin might be a useful vehicle for delivery of a toxic molecule to a target cell. Heterobifunctional antibodies have also been used as reagents in diagnostic assays. For example, Takashi and Fuller, Clin. Chem. 34 (1988), 1693, developed a sensitivity assay for human chorionic gonadotropin, and Gorog et al., J. Immun. Meth. 123, no. 131 (1989), developed a homogeneity assay for carcinoembryonic antigen (CEA) using heterobifunctional antibodies with a specific binding affinity for both an enzyme and an antigen. The heterobifunctional antibodies were used instead of an antibody-enzyme conjugate.

Methods for cloning and modifying the functional regions of abzymes and antigen-specific binding regions have been described. Iverson et al., Cold Spring Harbor Symp. 56, no. 273 (1989), described a method for constructing an expression vector for expression of abzymes and a method for expressing the Vl and Vh domains of an antibody.

Despite the progress in the development of new antibodies, further development is desirable. In particular, studies with conventional enzyme-antibody conjugates used in diagnostic assays show that the molecular size of the complex influences the diffusion coefficient in solution and therefore the speed of the associative immunological reaction.

Conventional enzyme-antibody conjugates are prepared by introducing reactive groups into both the antibody and the enzyme, typically 2-5 groups per molecule. Due to the number of reactive groups, multiple bonding and aggregate formation are possible. Although conditions are optimized to minimize aggregation, a variable number of high molecular weight compounds are always present. In addition, conventional enzyme-antibody conjugates are not readily prepared in an exact 1:1 ratio of catalytic functionalities and binding functionalities, thereby reducing the effectiveness of the conjugate performance.

Summary of the invention

The invention relates to antibodies which actually possess both a catalytic function and an antigen specificity. Each such antibody, which shall be referred to for this specification as a "heterobifunctional catalytic antibody (HCAb)", comprises a first F(ab) portion, the variable domain of which exhibits intrinsic catalytic activity for a substrate, and a second F(ab) portion, the variable domain of which, for an antigen different from the substrate. is immunologically specific. An advantage of such antibodies used as reagents in diagnostic immunoassays or as therapeutic molecules is their low molecular weight. An HCAb would be free of the excess molecular weight of an enzyme linked to it; furthermore, the molecular weight could be further reduced, for example during the generation of F(ab)₂ hybrids, and hereinafter the term "HCAb" refers to complete antibodies or active fragments thereof.

For example, a 1:1 conjugate of IgG (mol. wt. 160,000) and alkaline phosphatase (mol. wt. 140,000) has a molecular weight of about 300,000. An intact HCAb of the invention, which has the ability to react with a substrate in the same manner as alkaline phosphatase and the ability to bind to an antigen, would have a molecular weight of about 160,000. According to a preferred embodiment, the molecular weight of the HCAb can be further reduced by the creation of an F(ab)2 hybrid.

According to one embodiment of the invention, an HCAb is prepared directly, thereby bypassing the multi-step process required to prepare a standard enzyme-antibody conjugate. As discussed above, conventional enzyme-antibody conjugates are prepared using a process that results in aggregate formation in a variable amount. According to one embodiment of the invention, an HCAb is chemically prepared by creating bonds between sulfhydryl groups in two different Fab fragments, one Fab fragment having intrinsic catalytic properties and the other Fab fragment having the ability to specifically bind an antigen. Thus, conditions are highly favorable for the formation of an HCAb with a 1:1 ratio.

Another embodiment of this invention relates to methods for producing HCAbs by hybridizing a catalytic antibody (abzyme) with an antigen-specific antibody, which is preferably a monoclonal antibody. The products of this hybridization have a molecular weight corresponding to that of the starting molecules, while still possessing the activities of both starting molecules.

This invention includes the use of such HCAbs in sensitive immunoassays for the detection of a particular analyte.

According to a preferred embodiment of the invention, HCAbs of the invention are produced which have the ability to specifically bind to an analyte detected in an immunoassay and an intrinsic catalytic ability to act on a substrate to generate a detectable signal. The HCAbs of the invention can replace enzyme-antibody conjugates which are predominantly used in immunoassays. For example, HCAbs have a catalytic domain with a phosphatase activity and the ability to specifically bind to a tumor marker, an active substance such as theophylline, hormones such as TSH, T4 and progesterone. Vitamins such as B12, and any other analyte for which monoclonal antibodies are available.

This invention also relates to diagnostic kits for detecting an antigen in samples comprising a predetermined amount of an HCAb which specifically binds to the antigen in an amount related to the amount of antigen in the sample and the substrate to which the antibody exhibits catalytic activity, such that the converted substrate is detectable or measurable. The kit also preferably comprises a solid support to which the HCAb has been directly or indirectly bound.

According to another embodiment of the invention, an HCAb is prepared that specifically binds to a unique cellular tumor marker and has a catalytic activity that is toxic to that cell and is used in the detection of prodrug activation.

Short description of the characters

Figure 1 shows an expression system for producing a heterobifunctional, catalytic antibody according to the invention in a host cell.

Figure 2 shows a Sephacryl size exclusion column profile of the hybrid F(ab')₂ (MOPC 167 x 9-49, see Example 2 below).

Figure 3 shows the catalysis of p-nitrophenylcarbonylcholine by the heterobifunctional catalytic antibody of the invention when it is simultaneously associated with theophylline immobilized on particles (see Example 4 below).

Detailed description of the invention

HCAbs of the invention can be obtained by several different methods, such as the selective reduction and reassociation of two different antibody molecules or by hybridization of two different hybridomas. Genetic engineering techniques can also be used. The yield varies with the method and the starting materials used.

Unconventional methods, such as genetic engineering techniques, where the complementary pairing regions within the variable regions of an antibody molecule are engineered to produce HCAbs with a desired catalytic activity and with the ability to specifically bind to a desired antigen, are suitable to improve the annealing efficiency.

According to one embodiment, the HCAbs are produced by cell types engineered for this purpose. One cell type is a quadroma and is generated by the somatic cell fusion of two hybridomas, each parent hybridoma producing a monoclonal antibody, one of the monoclonal antibodies having a exhibits catalytic activity for a particular substrate and the other monoclonal antibody is specific for a particular antigen. Another type of cell is a trioma and is produced by the fusion of a hybridoma and a lymphocyte, each producing antibodies, with one type of antibody exhibiting catalytic activity and one antibody being specific for the antigen.

According to conventional methods, to produce a hybridoma that secretes catalytic antibodies capable of catalyzing a particular reaction, an antigen must first be produced that is structurally similar to an energetic class of intermediate expected in a chemical reaction. The antigen or hapten is often bound to a larger antigenic carrier molecule. Such antigens or haptens preferably comprise a transition state analog, such as a phosphonate ester in the formation of an amide or ester molecule.

Carrier molecules useful in the invention are well known and are generally proteins. They include tetanus toxoid or keyhole limpet hemocyanin, poly-L-lysine, thyroglobulin, albumins, erythrocytes and the like.

Methods for linking carrier molecules to the antigen are well known in the art (see Wong S.H. in Chemistry of Protein Conjugation and Cross-Linking, CRC Press (1991)). Bridging groups useful in the invention include m-maleimidobenzoyl-N-hydroxysuccinimide ester, glutaraldehyde, N-hydroxysuccinimidylglutaryl chloride and N-hydroxysuccinimidyladipoyl chloride.

The antigen-carrier conjugate is used to immunize animals to provide a population of spleen cells that produce the desired antibody. Immunization can be carried out by conventional in vivo methods. The splenocytes are then fused with cells from one of the commercially available myeloma cell lines, such as SP2/O, available from the ATCC. The supernatants from individual cultures are screened for an antibody response using a means such as a 96-well plate to which an antibody to immunoglobulin has been bound. The supernatants are incubated in the plates for a time sufficient for the antibody in the supernatants to bind to the plate. Subsequently, the substrate, which is cleaved by an antibody having the desired catalytic activity, is added to the plates and incubated. The plates are periodically examined for signs of catalytic activity. Positive clones are propagated and further evaluated for signs of hydrolytic or catalytic activity with the desired substrate.

Separately, a fusion partner is developed, which is either a hybridoma that produces monoclonal antibodies that specifically bind to a desired antigen or a spleen cell from an animal immunized with the desired antigen.

As mentioned above, HCAbs can be obtained by generating a trioma or quadroma. Monoclonal antibody-producing hybridomas are prepared in the usual manner.

Myeloma cell mutants lacking the enzyme hypoxantine-guanine phosphoribosyl transferase (HGPRT) are preferred. These cells do not grow in medium containing hypoxantine, aminopterin and thymidine (HAT). Spleen cells contain HGPRT but cannot divide continuously. Consequently, only hybridomas containing HGPRT (hybridomas fused with spleen cells, i.e. triomas) survive in HAT medium. Such cell lines are called HAT-resistant.

These hybridomas are generally backselected for HAT sensitivity before further fusion.

Backselection is performed by growing cells in increasing concentrations of 8-azaguanine (starting at 1 mg/ml, increasing to 20 mg/ml) in complete growth medium to enrich for HGPRT-deficient and mutant cells. HAT sensitivity can be generated in a similar manner using bromodeoxyuridine to induce thymidine kinase deficiency.

These back-selected antibody-producing hybridomas can then be directly fused with antibody-secreting spleen cells from an animal immunized with the antigen, creating a trioma. Such triomas can then be selected using HAT medium.

To produce a quadroma, one hybridoma line with HAT resistance is usually maintained and the other hybridoma line is back-selected for HAT sensitivity and, in some cases, further modified for resistance to biological inhibitors or cytotoxic agents. Ouabain has been most commonly used, but emetine, actinomycin and neomycin can also be used. In this way, when a HAT-sensitive/ouabain-resistant hybridoma is fused to a HAT-resistant/ouabain-sensitive hybridoma, quadromas are produced that are both HAT and ouabain resistant and are therefore easily identified.

A recombinant expression system can also be used to produce the HCAbs of the invention by culturing cells transformed by recombinant vectors. Heavy and light antibody chains can be re-engineered to provide selective complementarity as well as binding of the relevant genes to a single chromosome. Indeed, it is expected that a recombinant system would be a more stable system for the expression of HCAbs and would serve as a vehicle for the expression of complementarity-determining regions from the variable domains of other antibody specificities.

The HCAb molecule comprising each of the variable domains is a product of the annealing of two types of polypeptide chains (heavy and light) encoded by different genes. As is well known, the genes that normally encode the synthesis of each chain have a complex organization in which the required nucleic acid base pair sequences occur in a discontinuous, linear arrangement. That is, the genetic information required for the expression of the elements needed to construct the intact heavy or light chain is separated by intervening, unexpressed genetic information. The genetic information required for expression of the heavy chain gene includes promoter sequence information, a separate leader sequence, a short diversity sequence, a short junction sequence, a constant sequence, and termination codons; in comparison, the light chain gene arrangement includes promoter sequence information, junction and termination codons. The transcript mRNA (i.e. messenger ribonucleic acid) of the immunoglobulin gene consists of a linear, continuous array of complementary ribonucleic acid sequences. A recombinant expression vector encoding expression of the continuous sequence could be prepared by reverse transcription of the messenger ribonucleic acid into deoxyribonucleic acid. The expression vector is designed to contain selective markers and to provide for expression of both heavy and light chains within an HCAb. Furthermore, the genetic information is expressed by introducing the genes encoding immunoglobulin heavy and light chain synthesis into a suitable vehicle for expression, such as bacteria, fungi, insect or mammalian cell culture.

The present inventors designed a recombination scheme and provide a general description thereof. Briefly, the expression system illustrated in Figure 1 comprises a nucleotide base sequence for regulating gene expression. The phosphoglycerate promoter sequence has been used to regulate the synthesis of complex molecules, including immunoglobulin, and is known to those skilled in the art. The promoter (P) sequence is followed by a ribosome binding site (RB). A sequence containing a start codon (v.v. ATG) and a leader sequence would also be incorporated into the gene; the sequence TCT GCC CAA ATT is known to those skilled in the art of gene expression in yeast as a leader sequence capable of causing secretion of the polypeptide. The leader sequence is immediately followed by sequences encoding the variable (V) and constant (C) domains, respectively. This arrangement would include the proximal orientation of the information encoding the variable domain.

It is well known that within the variable domains of both the heavy chain and the light chain encoding gene sequences, hypervariable sequences of the genetic code exist. In addition, the hypervariable sequences encode complementary Peptide sequences or complementarity determining regions (vv CDR) within the antigen (hvs or hvs') or catalytic (HVS or HVS') binding clefts of the HCAb immunoglobulin molecule. Regions containing these hypervariable sequences are flanked at both ends by defined restriction enzyme cleavage sites. The value of such a design is that it would allow the removal and exchange of CDRs in blocks or "cassettes"; the latter term is generally used by those skilled in the art to designate the mobile blocks of genetic information. This design thus allows innumerable changes of the catalytic or antigen-specific sequences and also the introduction of sequences from biologically active molecules such as enzymes.

The variable domain is linked to the constant (C) domain of the light chain or the first constant (C1) domain of the heavy chain via short elements of the joining (J or DJ) sequence in a manner that results in the expression of the protein structure consistent with the known organization of the immunoglobulin. Some sequences can be extended or shortened in accordance with design requirements.

The pairing of immunoglobulin chains within the cell can occur randomly, with heavy chain and light chain pairing occurring without regard to their respective antigen specificities or catalytic specificities. Such misalignment can reduce the efficient recovery of functionally active HCAbs. As a further development, the genes encoding the first heavy chain constant domains (CL' or cl') and the light chain constant domain (CL or cl) are modified to selectively provide complementary sequences between the corresponding heavy and light chain pairs.

A sequence encoding a hinge (H) region is included, followed by all or part of the second heavy chain constant domain (C2). As with the pairing of the light and heavy chains, complementary sequences would be selectively introduced into the code for heavy chain synthesis to allow only the pairing of the catalytic (CH) antibody half with the antigen-specific (CH') antibody half. The sequence can then be followed by genetic information encoding the third constant region (C3), followed by a termination sequence.

The production of the HCAbs of the invention is not limited to in vivo expression systems. Biochemical methods are also used. One of them, which was used to produce an HCAb of the invention, involved limited proteolysis and reduction of two different antibodies, one monoclonal antibody being a catalytic antibody, e.g. MOPC 167 (readily available from the National Cancer Institute of the National Institutes of Health, Bethesda, MD. USA), and the other being any antibody with a specific Binding affinity for an analyte of interest to produce unifunctional F(ab') hybrids, followed by reannealing and reoxidation to generate interchain disulfide bridges. To promote the formation of HCAbs, the thiols of one antibody F(ab') fragment are activated with Ellman's reagent or with the bifunctional crosslinking agent o-phenylenedimaleimide prior to mixing with an equimolar amount of the second antibody F(ab') fragment containing free thiols.

The HCAbs of the invention obtained using any of the methods described above can be purified by affinity chromatography. The primary purification uses an immobilized transition state analog, with a secondary purification step occurring on an immobilized antigen column.

The purified HCAb can then be used for any diagnostic application, such as an immunoassay for analyte detection that requires an enzyme-conjugated antibody, such as sandwich assay arrangements, among others, and the present invention is not limited to any specific arrangements or protocols. Also, catalytic sites from existing enzymes, such as alkaline phosphatase, can be replaced with the complementarity determining regions within the catalytic antibody molecules.

As used herein, the term "solid phase support" refers to an insoluble material to which a component of the assay may be bound. Known materials of this type include hydrocarbon polymers such as polystyrene and polypropylene, glass, metals, and gels. Such supports may be in the form of beads, tubes, strips, disks, and the like. Magnetic particles are particularly preferred for use in the assays of the invention.

Such antibodies may also be used as pharmaceuticals and in methods of treating diseases involving target cells having antigenic determinants on the surface, wherein the HCAb catalytic moiety provides the therapeutic activity. Such diseases include cancers, non-malignant tumors, and the like, if the antibody is toxic to the diseased target cells.

The invention will be better understood by reference to the following, but non-limiting, examples.

Example 1: Heterobifunctional catalytic antibody (HCAb) suitable for ester-specific hydrolysis and specific for human IgG Production of catalytic antibody producing hybridomas

Eight-week-old female BALB/c mice are challenged on a two-week basis with p-aminophenylmethylphosphate linked to the tetanus toxoid via hemisuccimide in the para position in incomplete Freund’s adjuvant for immunized intraperitoneally. The mice are killed after 10 to 20 weeks by cervical dislocation and the spleens are surgically removed.

The spleens are transferred to a Petri dish containing 5 ml of Dulbecco's modified Eagle's medium (DME) with 100 µg/ml gentamicin. The splenocytes are released by plucking the tissue apart. The splenocytes are washed once with DME by centrifugation at 250 g for 5 minutes at 18-25 ºC and then combined in a ratio of 5:1 with washed SP2/O cells, a commercially available plasmacytoma cell line.

Cells are pelleted by centrifugation at 200 g for 6 minutes at 18-25ºC, excess supernatant is removed and cells are resuspended in 1 ml of buffered 50% polyethylene glycol solution. Cells are incubated at room temperature for 120 seconds before centrifugation at approximately 400 x g for 6 minutes. Cells are then centrifuged at 250 g for 5 minutes at 18-25ºC, gently washed with DME, resuspended in Iscove's modified DME with SP2/O conditioned medium and carefully transferred to a 75 cm2 flask.

After incubation overnight at 37ºC in a 5% CO2 atmosphere, the cells are resuspended in conditioned medium to 2 x 106 cells/ml and seeded into 96-well plates (250 µl/well). After 7-10 days, the supernatants are screened as follows. Flat-bottomed 96-well polystyrene plates are presensitized overnight with 100 mg/well goat anti-mouse IgG. The plates are washed and then the free binding sites of the plates are blocked with 1% bovine serum albumin (BSA) in 50 mM phosphate-buffered saline, pH 7.2, for 30 minutes, washed and dried. A 100 µl aliquot of the supernatant is added to each well.

The substrate, p-nitrophenyl acetate, is diluted in a neutral or phosphate-free buffer (15 mM Tris, 15 mM NaCl, pH 7.0), added to each well, and incubated for up to 24 hours. Plates are periodically examined for signs of ester hydrolysis, which is characterized by the appearance of free p-nitrophenol, a yellow chromogen. Positive clones are expanded and evaluated with a luminogenic luminescent substrate, such as AMPPD, for signs of hydrolytic activity, and the luminescent signal is measured.

Preparation of the merger partner

Murine hybridomas secreting monoclonal antibodies with antigen specificity to human IgG are subjected to selection guidelines to obtain a suitable fusion partner. First, cells lacking hypoxanthine phosphoribosyl transferase are identified as follows. 108 exponentially growing cells in Iscove's DMWM with 20% fetal bovine serum are harvested, washed and resuspended in medium containing 6.6 mM 8-azaguanine. The cells are cultured in presence of 8-azaguanine for approximately 24 hours or until approximately 80% of the cells have lost viability.

The cells are then washed and resuspended in the above culture medium, free of 8-azaguanine, and cultured until they reach a mass of about 10 g of cells, determined by cell counting. This procedure is repeated two more times. The cells are then subjected to a final selection in 8-azaguanine by culturing for 5 days in the presence of the inhibitor.

Verification of HGPRT⊃min; deficiency is confirmed by culturing in Iscove's medium containing HAT. The cells are then further selected for resistance to oubain. The selection procedure is the same as that described above for 8-azaguanine. Cells expressing an antibody to IgG and resistant to HGPRT⊃min; and oubain are selected.

Fusion of hybridomas

The hybridomas obtained according to the methods described previously are fused by combining them in a 1:5 hybridoma to target cell ratio in 1 ml of buffered 50% polyethylene glycol solution. The cells are incubated at room temperature for 120 seconds before centrifugation at approximately 400 x g for 6 minutes. All cells are centrifuged, gently rinsed with DME, resuspended in Iscove's modified DME with SP2/O conditioned medium and carefully transferred to a 75 cm2 flask.

After incubation overnight at 37ºC and 5% CO2, cells are diluted to 2 x 106 cells/ml in conditioned medium and seeded into 96-well plates (250 µl/well). After 7-10 days, supernatants are screened using flat-bottomed 96-well polystyrene plates presensitized overnight with human IgG. Plates are washed and then the free binding sites of the plates are blocked with 1% bovine serum albumin (BSA) in 50 mM phosphate buffered saline, pH 7.2, for 30 minutes, washed and dried. A 100 µl aliquot of the supernatant is added to each well and the substrate, p-nitrophenyl acetate diluted in neutral or slightly acidic phosphate-free buffer, is added prior to incubation for up to 24 hours. Plates are examined periodically for signs of ester hydrolysis, which is characterized by the appearance of free p-nitrophenol, a yellow chromogen. Positive clones are expanded and evaluated with a luminescent substrate for signs of hydrolytic activity. Wells showing substrate conversion are determined to contain hybriddones that both recognize the antigen (i.e., in this case, human IgG) and cleave the distinct substrate.

Example 2: HCAb with phosphorylcholine esterase activity and an affinity for human IgG

MOPC 167, a commonly available clone with a reported phosphorylcholine carbonate cleavage activity described by Schultz et al. in Science 234 (1986), 1570, is fused with hybridomas producing antibodies specific for human IgGs to yield a quadroma.

The IgG-specific clone is back-selected for shared 8-azaguanine resistance and ouabain sensitivity according to the procedures described in Example 1. Cell fusion is performed as previously described. After plating and growing in culture for several days, cell supernatants are screened using 96-well plates sensitized with purified human IgG. Culture supernatants are transferred to the plates, incubated for a specified period of time, and washed with a buffered solution.

S-acetylthiocholine (i.e., the substrate), which can be cleaved by the catalytic antibody produced by MOPC 167, is added to the plates. Wells showing substrate conversion are determined to contain an HCAb.

Example 3: HCAb with phosphorylcholine esterase activity and an affinity for theophylline

Clone 9-49 is a hybridoma producing a monoclonal antibody against theophylline and is available from Sanofi Diagnostics Pasteur (Chaska, MN 55318, USA).

Any theophylline-specific monoclonal antibody can be used in this example instead of monoclonal antibody 9-49.

Hybridoma cells of clone 9-49 are fused with cells of clone MOPC 167. The cell fusion is carried out as previously described.

Selection of fused cells, characterization of catalytic activity and theophylline binding capacity of the newly generated HCAb (i.e., MOPC 167 x 9-49) is performed exactly as in Example 2, except that theophylline-presensitized microplates (or other solid supports) are used instead of goat anti-mouse IgG-presensitized plates.

Example 4: Biochemical preparation of an HCAb with phosphorylcholine esterase activity and theophylline affinity

Production of antibodies

Both MOPC 167 and 9-49 were propagated in sensitized BALB/c mice.

The MOPC 167 antibody was isolated from ascites fluid by immunoaffinity chromatography on Affigel G 10 with bound p-aminophenylphosphorylcholine and eluted with mild Pierce elution buffer.

Separately, monoclonal antibody 9-49 was isolated by successive fractionation using a DE-S2 Whatman column and 0.3 M sodium phosphate buffer, pH 7.2, followed by size separation on a Sephacryl G-200 column equilibrated with 0.15 M PBS, pH 7.4.

Production of Fab fragments

Each of the above antibody preparations was dialyzed overnight in 100 mM sodium acetate buffer, pH 4.3, at 4ºC. To determine the most appropriate enzymatic cleavage procedure, each antibody was then contacted with porcine gastric pepsin in a ratio of 100:1 at 37ºC, samples taken at several time points, and the cleavage products analyzed by SDS-PAGE gel electrophoresis using a Pharmacia PHAST system. The optimal conversion time for MOPC 167 was 4 hours or 2 hours for 9-49.

Antibody preparations were reacted according to the optimal incubation times observed and the reaction was stopped by raising the pH to 8. The cleavage products were then separated in 200 mM Tris-HCl, pH 8.0, with 10 mM EDTA over a Biocryl matrix on a Pharmacia FPLC system at a flow rate of 1 ml/min. The resulting F(ab')2 fraction of each antibody was further reduced by the addition of 2-mercaptoethanol to a final concentration of 20 mM and incubated for 30 minutes at 37°C; all subsequent reaction steps were carried out at 4°C. Each resulting Fab' fragment was separated from excess 2-mercaptoethanol by passage through a Sephadex G-25 column equilibrated with 50 mM sodium acetate, pH 5.3, containing 0.5 mM EDTA.

Two volumes of the 9-49 Fab' fraction were slowly combined with one volume of 12 mM o-phenylenedimaleimide dissolved in anhydrous dimethylformamide. After approximately 30 minutes, the reactants were passed over a Sephadex G-25 column equilibrated with 50 mM sodium acetate, pH 5.3, containing 0.5 mM EDTA.

Production of MOPC 167 x 9-49-HCAb

The modified 9-49 Fab' fragment was combined with the MOPC 1 67 Fab' fragment in a ratio of 5:1, concentrated to a total protein concentration of approximately 5 mg/ml and incubated overnight at approximately 4-5°C. The resulting reactants were filtered through a 0.22 µm membrane and purified by size exclusion chromatography. on a Sephacryl® column in Tris-HCl buffer, pH 8.0, with EDTA (see Figure 2). The resulting bifunctional antibody was eluted from the size exclusion column with an average apparent molecular weight of approximately 120 kD (the first vertical bar in Figure 2), as determined with molecular weight standards. This material was used in the inventors' subsequent studies.

Example 5: HCAb assays Substrate production

p-Nitrophenylcarbonylcholine was prepared from p-nitrophenylchloroformate and choline under anhydrous conditions according to the method of Shokat et al. in Ann. Rev. Immunol. 8 (1990), 335. The reaction product was rinsed thoroughly with dry pyridine. After removal of the solvent, the reaction product was found to have a melting temperature of 158ºC and to decompose to a yellow liquid at 160ºC. This material was further subjected to analysis by mass and infrared spectroscopy.

Enzyme immunoassay for the detection of the bifunctional antibody

MOPC 167, 9-49 and MOPC 167 x 9-49 antibody fractions and normal mouse IgG were separately diluted to 10 µg/ml in 150 mM Tris, pH 8.0, containing 1% bovine serum albumin. A 1 ml aliquot of each diluted antibody was combined with 1 mg of washed magnetic particles (Rhone Poulenc) sensitized with p-aminophenylphosphorylcholine. After 1 hour, the particles were washed with the binding buffer and combined with 1 µg of theophylline-labeled alkaline phosphatase (1:3 ratio). After an incubation period of 15 minutes at room temperature, the particles were washed with 100 mM Tris (pH 8.0) containing 0.05% FC-100, a surfactant available from 3 M Corporation, St. Paul, USA. Lumiphos, available from Lumigen (USA), was added to each tube and the signal measured in a Berthold luminometer.

Immunoassay detection of the catalytic conversion of p-nitrophenylcarbonylcholine

100 µg of MOPC 167, 9-49, MOPC 167 x 9-49 or normal murine IgG were mixed separately with 1 mg of either theophylline-sensitized or goat immunoglobulin-coated magnetic particles for 2 hours. The particles were then washed with MOPC buffer (15 mM Tris, pH 7.0, containing 15 mM sodium chloride). p-Nitrophenylcarbonylcholine was dissolved in MOPC buffer to a concentration of 25 mM. The substrate solution was added to the vortexed particles and incubated at 29.3ºC until measurement (see Figure 3); 30 seconds Before each measurement, the particles were magnetically separated and the absorbance was read at 410 μm in a Hewlett Packard diode array spectrophotometer.

Fractions, including the major portion containing the 120 kD molecule and lower molecular weight molecules, were tested for the ability to simultaneously bind theophylline and a choline derivative using the immunoassay procedures described above. The results showed that fractions averaging 120 kD molecular weight were able to retain 85 to 100% of an enzyme-labeled theophylline conjugate on particles sensitized with the choline derivative, while the next lower molecular weight pool retained only one-third of the total added conjugate activity. This observation is expected to reflect dilution of the intact bifunctional antibody and an increase in monofunctional Fab' fragments.

In another experiment, the bifunctional antibody molecule was bound in the reverse orientation to the theophylline derivative bound to the solid phase. Figure 3 illustrates the rate increase in substrate conversion associated with the system containing the bifunctional antibody, and that no increase is seen with the monofunctional Fab' fragments included in the assay. No increase in substrate conversion was observed when a solid phase lacking the theophylline derivative was used (not shown); this observation was interpreted as confirming the hypothesis that in this type of assay, both the antigen target and the substrate conversion activities must be combined within the same molecule.

Thus, the antibody product of the present inventors with a molecular weight of about 120 kD showed dual recognition capacity for theophylline and choline derivatives (Figure 2) as well as the catalytic conversion of p-nitrophenylcarbonylcholine when simultaneously associated with theophylline immobilized on particles (Figure 3). This coupling is unusual in that the starting antibodies were of the A(K) and G1(k) isotypes, respectively.

Competition assay for an antigen

In performing the assay, 200 µl of antigen-containing samples or standards in 10 mM Tris with 5 µm KCN, pH 8.0, are transferred to 10 x 60 mm borosilicate glass test tubes. To each reaction tube is added 100 µl (0.1 µg) of the HCAb, which specifically binds the antigen and which catalyzes a reaction with a specific substrate. The reactants are incubated for 15 minutes at 37ºC in a shaking water bath with stirring at approximately 60 rpm. 50 µl of the particles are added to the reaction mixture and incubated with stirring for an additional period of 30 minutes at 37ºC. The reaction is stopped by sedimentation for 2 minutes in a magnetic field. The tubes are washed twice with 1 ml of 50 mM Tris, 0.15 M NaCl, 0.1% Triton X-100, pH 7.4. 200 µl of substrate is added to each tube and the reaction is initiated by a static incubation at 37ºC for 5 minutes. The tubes are then cooled to room temperature for 2 minutes and the resulting signal is detected using a spectrophotometer or a luminometer, depending on the substrate used.

Example 6: Manufacture of a medicinal product

HCAbs of the invention can be used in the treatment of certain diseases with cell-bound targets. An HCAb can be prepared using the methods described above, wherein a portion of the antibody molecule recognizes a unique cellular marker (i.e., carcinoembryonic antigen (CEA)) and a portion of the antibody molecule has toxic diphosphoribotransferase/NAD-transferase activity (diphtheria toxin) or activates a potentially toxic compound.

The drug is injected into a patient with the disease and the antibody molecules bind to cells bearing the marker. Until the antibody is internalized, the catalytic half of the molecule specifically transfers nicotine adenine dinucleotides to elongation factor 2, after which certain key metabolic processes cease to function, resulting in the death of the target cell.

Claims (10)

1. A heterobifunctional catalytic antibody useful in a diagnostic assay in the manner of an enzyme-labeled antibody, the antibody comprising a constant region and two Fab fragments, the variable domain of the first Fab fragment exhibiting intrinsic catalytic activity for a substrate and the variable domain of the second Fab fragment having antigen binding specificity for an antigen, the antigen being different from the substrate, such that the catalytic activity for the substrate can be correlated with the binding of the antigen. 2. Heterobifunctional catalytic antibody according to claim 1, wherein the constant region consists of two constant domains which are respectively linked to each of the variable domains. 3. Heterobifunctional catalytic antibody according to claim 1, wherein the catalytic activity of the antibody is that of a phosphatase. 4. Heterobifunctional catalytic antibody according to claim 1 or 2, wherein the catalytic part of the antibody is derived from the MOPC 167 antibody. 5. A process for producing a heterobifunctional, catalytic antibody according to claims 1 to 4, comprising: a) fusing hybridomas which produce monoclonal catalytic antibodies with either lymphocytes or hybridomas which produce antibodies against the desired antigen or hapten, b) culturing the hybrid cells obtained in a), c) Selecting the clones which produce antibodies directed against the antigen or hapten and which exhibit catalytic activity. 6. A process for producing a heterobifunctional, catalytic antibody according to claims 1 to 4, comprising: (a) Transfection into a cell or microorganism of a recombinant expression vector whose DNA encodes the heavy and light chains of the antibody, so that the first variable region of the antibody has the desired antigen or hapten specificity and the second variable region has the desired has catalytic properties, b) allowing the cells or microorganisms to grow in such a way that the antibody is expressed, c) Collecting the antibody. 7. The method of claim 6, wherein the DNA region encoding the antibody constant region comprises a sequence that enforces subsequent annealing of the necessary light chain/heavy chain and heavy chain/heavy chain combinations. 8. A process for producing a heterobifunctional, catalytic antibody according to claims 1 to 4, comprising a) reducing the disulfide bonds linking the two heavy chains of a catalytic antibody and those of an antigen-binding antibody, b) combining the two types of resulting chains and reacting them with a cross-linking agent capable of rearranging antibody heavy chains, and c) Obtaining rearranged antibodies that possess catalytic properties coupled with a affinity for the antigen. 9. A method for testing for an analyte in a sample, comprising: a) contacting the sample with a known amount of a heterobifunctional, catalytic antibody according to claim 1 to allow the antibody to bind the analyte present in the sample: b) separating bound antibody from unbound antibody; c) measuring the extent of catalytic activity associated with either the analyte-antibody complex or the amount of unbound antibody by mixing them with the substrate to determine the amount of analyte present. 10. An immunoassay kit comprising: - a heterobifunctional, catalytic antibody according to claims 1-4, - a substrate of the catalytic antibody, - a means or device for determining the transformation of the substrate.